Neural Waveform Distinguishment via Gradient Encoder Ensemble
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current neural waveform tests struggle to accurately distinguish multiple intermingled neural waveforms due to the large size of electrodes relative to neurons, leading to difficulties in separating and analyzing brain activity signals effectively.
Innovation Solution
A neural waveform distinguishing apparatus and method that utilize a learning-based encoder ensemble to extract features from gradient waveforms by calculating pointwise slopes, concatenate codes from multiple encoders, and apply clustering techniques like DBSCAN to separate neural waveforms based on extracted features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If electrodes are used to sense neural signals, then neural waveforms can be detected, but the large size of electrodes relative to neurons causes multiple neural waveforms to be sensed simultaneously and intermingled, making accurate separation difficult
Solution Approach 1:
The patent applies segmentation by dividing the complex waveform separation task into multiple processing stages: gradient calculation, feature extraction using encoder ensemble, and clustering. This multi-stage approach breaks down the difficult problem of separating intermingled neural waveforms into manageable steps, improving separation accuracy without requiring physically smaller electrodes
Solution Approach 2:
The patent introduces gradient waveforms as an intermediary representation between the raw sensed neural signals and the final separated waveforms. By calculating pointwise slopes to create gradient waveforms, the system creates an intermediate form that enhances distinguishability between different neural sources, making the subsequent clustering more effective
2Quantity of substance
If threshold-based filtering is used to extract neural signals with intensity greater than or equal to a pre-designated threshold, then some neural waveforms can be obtained, but only two to ten waveforms are detected intermingled, limiting in-depth analysis
Solution Approach 1:
The patent transforms the waveforms by calculating their gradients (pointwise slopes), changing the parameter representation from raw amplitude values to rate-of-change values. This parameter transformation enhances the distinguishability between different neural waveforms, allowing more waveforms to be accurately separated even when they appear intermingled in the original signal
Solution Approach 2:
The patent uses an encoder ensemble that combines multiple encoders with different numbers of hidden layers to extract features. This composite approach, analogous to composite materials, leverages the strengths of different encoder configurations to achieve better waveform separation than any single encoder could achieve alone, enabling accurate distinction of more neural waveforms
3Measurement precision
If multiple encoders with different numbers of hidden layers are used to extract features, then more comprehensive feature ensembles can be obtained, but the device complexity and computation increase
Solution Approach 1:
The encoder ensemble segments the feature extraction task across multiple encoders with different architectural complexities (different numbers of hidden layers). Each encoder contributes features at different levels of abstraction, and their outputs are concatenated to form comprehensive feature ensembles. This segmentation allows the system to achieve high extraction accuracy while managing complexity through modular design
Solution Approach 2:
The patent adds dimensionality by combining features from multiple encoders with different hidden layer configurations. This creates a multi-dimensional feature space that captures diverse characteristics of the neural waveforms, improving discrimination accuracy. The concatenation of codes from multiple encoders effectively stacks features across different architectural dimensions
Data Source
AI summary
A neural waveform distinguishment apparatus includes: a neural waveform obtainment unit that obtains multiple neural waveforms in a pre-designated manner from neural signals sensed by way of at least one electrode; a preprocessing unit that obtains multiple gradient waveforms by calculating pointwise slopes in each of the neural waveforms; a feature extraction unit comprising an encoder ensemble composed of multiple encoders, which have a pattern estimation method learned beforehand and include different numbers of hidden layers, where the feature extraction unit obtains multiple codes as multiple features extracted by the encoders respectively from the gradient waveforms and concatenates the codes extracted by the encoders respectively to extract a feature ensemble for each of the gradient waveforms; and a clustering unit that distinguishes the neural waveforms corresponding respectively to the gradient waveforms by clustering the feature ensembles extracted respectively in correspondence to the gradient waveforms according to a pre-designated clustering technique.